U.S. patent application number 13/380796 was filed with the patent office on 2012-11-01 for method for determining the susceptibility of a cell strain to drugs.
This patent application is currently assigned to ASSISTANCE PUBLIQUE - HOPITAUX DE PARIS. Invention is credited to Alexandre Alanio, Jean Y. Brossas, Annick Datry, Jean L. Golmard, Carine Marinach-Patrice, Dominique Mazier, Martine Palous.
Application Number | 20120276577 13/380796 |
Document ID | / |
Family ID | 41059543 |
Filed Date | 2012-11-01 |
United States Patent
Application |
20120276577 |
Kind Code |
A1 |
Mazier; Dominique ; et
al. |
November 1, 2012 |
Method for Determining the Susceptibility of a Cell Strain to
Drugs
Abstract
The present invention relates to a method for determining the
susceptibility of a cell strain to a compound intended for
controlling the growth of said cell strain, comprising: growing the
cell strain in a first compound-free culture medium and in at least
a second culture medium comprising the compound at a test
concentration; obtaining mass-spectrometry spectra for a protein
extract of the cell strain grown in the first culture medium and
for a protein extract of the cell strain grown in the second
culture medium; comparing the mass spectrometry spectra; deducing
that the cell strain is sensitive to the compound at the test
concentration if the mass spectrometry spectra are significantly
different.
Inventors: |
Mazier; Dominique; (Paris,
FR) ; Marinach-Patrice; Carine; (Juvisy Sur Orge,
FR) ; Alanio; Alexandre; (Paris, FR) ;
Golmard; Jean L.; (Paris, FR) ; Palous; Martine;
(Fontainebleau, FR) ; Datry; Annick; (Paris,
FR) ; Brossas; Jean Y.; (Trainou, FR) |
Assignee: |
ASSISTANCE PUBLIQUE - HOPITAUX DE
PARIS
Paris
FR
UNIVERSITE PIERRE ET MARIE CURIE (PARIS 6)
Paris
FR
|
Family ID: |
41059543 |
Appl. No.: |
13/380796 |
Filed: |
June 25, 2010 |
PCT Filed: |
June 25, 2010 |
PCT NO: |
PCT/EP2010/059101 |
371 Date: |
March 28, 2012 |
Current U.S.
Class: |
435/32 ;
435/288.7 |
Current CPC
Class: |
G01N 2560/00 20130101;
C12Q 1/025 20130101; G01N 33/5008 20130101; G01N 33/6848
20130101 |
Class at
Publication: |
435/32 ;
435/288.7 |
International
Class: |
C12Q 1/18 20060101
C12Q001/18; C12M 1/34 20060101 C12M001/34 |
Foreign Application Data
Date |
Code |
Application Number |
Jun 25, 2009 |
EP |
09305601.8 |
Claims
1. A method for determining the susceptibility of a fungus strain
to a compound intended for controlling the growth of said fungus
strain, comprising: growing the fungus strain in a first
compound-free culture medium and in at least a second culture
medium comprising the compound at a test concentration; obtaining
mass-spectrometry spectra for a protein extract of the fungus
strain grown in the first culture medium and for a protein extract
of the fungus strain grown in the second culture medium; comparing
the mass spectrometry spectra; deducing that the fungus strain is
susceptible to the compound at the test concentration if the mass
spectrometry spectra are significantly different.
2. The method according to claim 1, wherein the fungus strain is a
Candida strain.
3. The method according to claim 1, wherein the fungus strain is a
Candida strain selected from a Candida albicans strain, a Candida
glabrata strain, a Candida tropicalis strain, a Candida
parapsilosis strain, a Candida kefyr strain, a Candida krusei
strain, a Candida dubliniensis strain, a Candida guillermondii
strain and a Candida lusitaniae strain.
4. The method according to claim 1, wherein the fungus strain is a
Candida albicans strain.
5. The method according to claim 1, wherein the compound is an
antifungal compound.
6. The method according to claim 1, wherein the compound is an
antifungal azole compound selected from the group constituted of
fluconazole, ketoconazole, itraconazole, voriconazole,
posaconazole, isavuconazole, and ravuconazole.
7. The method according to claim 1, wherein the compound is
fluconazole.
8. The method according to claim 1, wherein mass spectrometry is
carried out by MALDI-TOF.
9. The method according to claim 1, wherein the protein extract is
obtained by an ethanol treatment of grown fungus cells followed by
a treatment with a mixture of formic acid and acetonitrile.
10. The method according to claim 1, wherein the culture medium is
a minimum medium.
11. The method according to claim 1, wherein the fungus strain is
grown during 16 hours.
12. The method according to claim 1, wherein the fungus strain is
grown in several culture media with increasing test concentrations
of the compound, and the minimal concentration of the compound
yielding a mass-spectrometry spectrum detectably different from the
mass-spectrometry spectrum obtained from the cell-extract of the
fungus strain grown in the compound free culture medium is
determined.
13. The method according to claim 12, wherein determining the
minimal concentration comprises: obtaining a mass spectrometry
spectrum from a protein extract of the fungus strain grown in the
culture medium having the highest test concentration of the several
culture media; comparing the mass spectrometry spectra obtained
from protein extracts of the several culture media to (i) the mass
spectrometry spectrum from a protein extract of the fungus strain
grown in the culture medium having the highest test concentration
and (ii) the mass spectrometry spectrum from the compound-free
culture medium; selecting the mass spectrometry spectrum obtained
from a protein extract of the several culture media with the lowest
compound concentration and which presents more similarity with the
mass spectrometry spectrum (i) than with the mass spectrometry
spectrum (ii); the lowest compound concentration being the minimal
concentration to be determined.
14. A device for performing a method wherein a fungus strain is
grown in several culture media with increasing test concentrations
of a compound intended for controlling the growth of said fungus
strain, and the minimal concentration of said compound yielding a
mass-spectrometry spectrum detectably different from the
mass-spectrometry spectrum obtained from the cell-extract of the
fungus strain grown in a compound-free culture medium is
determined, comprising: an obtaining device obtaining a
mass-spectrometry spectrum for a protein extract of said fungus
strain grown in a first compound-free culture medium and in several
culture media with increasing test concentrations of said compound;
a comparing device comparing the mass spectrometry spectra obtained
for protein extracts of the several culture media with increasing
test concentrations of said compound to (i) the mass spectrometry
spectrum of a protein extract of said fungus strain grown in the
culture medium having the highest assayed test concentration and
(ii) the mass spectrometry spectrum of a protein extract of said
fungus strain grown in the compound-free culture medium; a
determination device determining the minimal concentration by
determining the mass spectrometry spectrum obtained from a protein
extract of the several culture media with the lowest compound
concentration and which presents more similarity with the mass
spectrometry spectrum (i) than with the mass spectrometry spectrum
(ii); the lowest compound concentration being the minimal
concentration.
Description
FIELD OF THE INVENTION
[0001] The present invention relates to a method for determining
the susceptibility of a cell strain to a compound intended for
controlling the growth of said cell strain.
TECHNICAL BACKGROUND
[0002] Testing the susceptibility of cell strains, such as tumour
cell strains or microorganism strains, to drugs is a crude
challenge, in particular in view of the increasing prevalence of
drug resistance.
[0003] Indeed, besides cancer cells, resistances to drugs have
notably been evidenced in bacteria, protozoan parasites, and fungi,
such as yeasts and filamentous fungi. This has notably been
evidenced in the case of the determination of the susceptibility of
Candida albicans to fluconazole (FCZ). Candida albicans is the
leading cause of invasive candidiasis, a major hospital-acquired
infection. Fluconazole, an azole derivative agent, is one of the
main first-line therapy alternatives. Azole resistant strains have
emerged, possibly as a consequence of the use of azole-based
antifungal agents in iterative and long-term therapies.
[0004] In vitro susceptibility testing is essential both for
epidemiologic surveillance, e.g. to detect the emergence of
resistant-microorganisms, and to adapt therapy for a given
patient.
[0005] Susceptibility of cell strains to drugs is usually
determined following the well-known broth microdilution methods as
gold standards tests. These methods are based on growth inhibition
and involve the determination of the Minimal Inhibitory
Concentration (MIC). Such methods are notably recommended by the
European Committee on Antibiotic Susceptibility Testing (EUCAST)
and the Clinical Laboratory Standards Institute (CLSI)
(Rodriguez-Tudela. et al. J Clin Microbiol 45, 109-111 (2007);
Espinel-Ingroff et al. J Clin Microbiol 43, 3884-3889 (2005)).
[0006] For the EUCAST methodology, the MIC endpoint, for example
for fluconazole susceptibility testing, is determined as the drug
concentration inducing a 50% growth inhibition (IC.sub.50) with
respect to the control as measured after 24 h of growth with a
spectrophotometer. For the CLSI methodology, MIC endpoints are
defined visually as the point at which there is prominent reduction
in growth in the sample as compared to the control after 48 h of
incubation. This visual end-point correlates with 50% growth
inhibition (Rex at al. Clin Microbiol Rev 14, 643-658 (2001)).
[0007] Both reference methods are robust and reliable, though they
remain time-consuming and thus inadequate for routine determination
in an hospital context (Revankar et al. J Clin Microbiol 36,
153-156 (1998); Lass-Florl at al. Antimicrob Agents Chemother 52,
3637-3641 (2008)).
[0008] Thus, to circumvent this drawback, some commercial assays
such as the E-Test (AB-Biodisk) or YeastOne Panel (Trek Diagnostic)
have been proposed that can be used widely and easily in clinical
microbiology labs. Overall, they have been favourably compared with
the reference methods, while results with some pairs of
microorganism-drugs do not exactly correlate with the results of
the standards. Moreover, they still remain quite long to carry out
and reading of the assays may be particularly difficult. This is
particularly the case when testing C. albicans isolates against
fluconazole, since it frequently leads to a trailing phenomenon,
defined by a low-level growing of the colonies even over increasing
concentrations of the drug. More recently, this so-called
paradoxical effect has also been described for some Candida
isolates when tested against ecchinocandin drugs.
[0009] The patent application US 2008/0009029 describes a method of
determination of bacterial resistance to the ampicillin antibiotic.
To measure the bacterial resistance to antibiotics, the protein
profiles of bacteria are measured after cultivation in media
containing the antibiotics. However, the teaching of US
2008/0009029 is limited to the measurement of microbial (bacterial)
growth in the presence of antibiotics. This patent does not give
any insight into the possible measurement of fungal growth in the
presence of antifungal drugs. It does not either teach the
determination of the minimal concentration of drug inducing a
detectable change in mass spectrometry spectra.
[0010] Given these limitations, there is a clear need for the
development of an equally robust method for determining the
susceptibility of a fungus such as a yeast to drugs, with faster
turn-around times and where endpoints determination is
objective.
[0011] There is also a need in the art for a method for quantifying
the resistance of a cell strain to a drug, e.g. determining the
minimal concentration of drug inducing a detectable change in mass
spectrometry spectra.
[0012] It is therefore an object of the present invention to
provide such a method.
DESCRIPTION OF THE INVENTION
[0013] The present invention arises from the unexpected finding, by
the inventors, that the protein composition of a C. albicans strain
changes reproducibly in response to a particular drug concentration
to which it is subjected, and that this variation in protein
composition can be evidenced by mass spectrometry. Besides, the
inventors have also shown that the values obtained for the minimal
concentration of drug inducing a detectable change in mass
spectrometry spectra of a protein extract of C. albicans are
approximately equal (by two dilutions) with the minimal inhibitory
concentrations determined for C. albicans using a standard method
(CLSI).
[0014] The present invention thus relates to a method for
determining the susceptibility of a cell strain to a compound
intended for controlling the growth of said cell strain,
comprising:
[0015] growing the cell strain in a first compound-free culture
medium and in at least a second culture medium comprising the
compound at a test concentration;
[0016] obtaining mass-spectrometry spectra for a protein extract of
the cell strain grown in the first culture medium and for a protein
extract of the cell strain grown in the second culture medium;
[0017] comparing the mass spectrometry spectra; [0018] deducing
that the cell strain is sensitive to the compound at the test
concentration if the mass spectrometry spectra are delectably
different.
[0019] As intended herein, the expression "cell strain" relates to
any kind of eukaryotic or prokaryotic cell strain. In particular,
where the cell strain is an eukaryotic cell strain it can be from a
pluricellular or an unicellular organism. As will be clear to one
of skill in the art, the unicellular organism can notably such that
it develops into a pluricellular organism. Preferably, the cell
strain is a tumour cell strain or a microorganism strain.
Preferably, the cell strain is a microorganism strain selected from
the group constituted of a bacterial strain, a fungus strain, such
as a filamentous fungus, in particular of the Ascomycota (e.g. of
the Aspergillus or Fusarium genus) and Zygomycota phyla, or a yeast
strain, in particular of the Ascomycota and Basiodiomycota phyla, a
protozoan strain, and an algae strain.
[0020] In a most preferred embodiment, the cell strain is a yeast
strain, in particular selected from group consisting of a Candida
strain, a Saccharomyces strain, a Debatyomyces strain, a Pichia
strain, a Geotrichum strain, a Cryptococcus strain, a
Fisiobasidiella strain, and a Trichosporon strain. Most preferably
the cell strain is a Candida strain. Besides, among yeasts of the
Candida genus, it is preferred that the cell strain is a Candida
strain selected from a Candida albicans strain, a Candida glabrala
strain, a Candida tropicalis strain, a Candida parapsilosis strain,
a Candida kefyr strain, a Candida krusei strain, a Candida
dubliniensis strain, a Candida guillermondii strain and a Candida
lusitaniae strain, and particularly preferred that the cell strain
is a Candida albicans strain.
[0021] As intended herein, the term "compound for controlling the
growth of said cell strain" relates to a compound liable to kill
cells of the cell strain or to inhibit, partially or totally, the
growth of cells of the cell strain. Thus, the compound may notably
be an anti-tumour compound, an antibiotic or antibacterial
compound, or an antifungal compound. However, it is preferred that
the compound is an antifungal compound selected from an azole
compound, an echinochandin compound, such as caspofungin,
micafungin, or anidulafungin, a polyene compound, such as
amphotericin B or nystatin, and anti-metabolites, such as
flucytosine. Preferably, the antifungal compound is an azole
compound selected from the group constituted of fluconazole,
voriconazole, posaconazole, isavuconazole, ravuconazole,
ketoconazole, and itraconazole. Most preferably the compound is
fluconazole.
[0022] As intended herein the expression "determining the
susceptibility of a cell strain" relates to determining whether,
and within what measure, the compound as defined above kills or
inhibits the growth of cells of the cell strain. In particular,
"determining the susceptibility of a cell strain" relates to
determining the minimal concentration of the compound which yields
a detectable difference in the mass spectrometry spectra.
[0023] As intended herein, "mass spectrometry" relates to any
method enabling determining the m/z ratio of one or more molecules,
such as proteins, within a sample, such as a protein extract as
defined above, wherein m represents the mass and z the charge of
said molecules. Mass spectrometry as defined above can be carried
out by any one of the numerous mass spectrometry methods known in
the art, such as Matrix-Assisted Laser Desorption/Ionisation
Time-Of-Flight (MALDI-TOF) mass spectrometry, or Surface-Enhanced
Laser Desorption/Ionisation Time-Of-Flight (SELDI-TOF). However, it
preferred that mass spectrometry as defined above is carried out by
MALDI-TOF.
[0024] The expression "mass spectrometry spectra" relate to
recordings of the m/z ratios and optionally the quantities of the
various molecules, in particular proteins, contained in the protein
extracts submitted to mass spectrometry. Usually, mass spectrometry
spectra are graphs representing signal intensity (corresponding to
the quantity of molecule) as a function of the m/z ratio. The
association of signal intensity to an m/z ratio defines a peak. As
will be clear to one of skill in the art "a mass spectrometry
spectrum" as intended herein can be either obtained from one
recording or be the mean of a plurality of recordings.
[0025] As intended herein, "mass spectrometry spectra are
detectably different" in particular if a detectable difference in
intensity of m/z ratio can be established. The man skilled in the
art knows how to establish that two spectra present detectable
differences, in particular using exact permutation tests based on
Spearman rank correlation coefficients, such as described in
"Design and Analysis of DNA Microarray Investigations", R M Simon
et al, SPRINGER, 2003, in particular on pages 68 and 123.
[0026] Numerous procedures are known in the art for extracting
proteins from cells and one of skill in the art knows how to adapt
them depending on the type of cell strain. Accordingly, any one of
these extraction methods can be used in the method of the
invention. However, it is preferred, within the frame of the method
according to the invention, that the protein extract is obtained by
an ethanol treatment of grown cells followed by a treatment with a
mixture of formic acid and acetonitrile, in particular where the
cell strain is a yeast strain, more particularly a Candida
strain.
[0027] As intended herein, the expression "culture medium" relates
to any medium liable to sustain the growth of cells of the cell
strain. Preferably, the culture medium as defined above is a
minimum medium. Preferably also, the medium is a liquid medium. As
will be clear to one of skill in the art, the composition of the
compound-free culture medium and the culture medium comprising the
compound at a test concentration should preferably identical except
for said compound.
[0028] The cell strain can be grown for any amount of time provided
it is sufficient for the compound to induce significant changes in
the protein content of the cell strain. However, so that the method
is carried out as quickly as possible, it is preferred that the
amount of time for growing the cells is the minimal time for the
compound to induce significant changes in the protein content of
the cell strain. Thus, the cell strain is grown during less than 24
hours, more preferably during less than 20 hours, and most
preferably during about 15 hours.
[0029] In an embodiment of the above-defined method, the cell
strain is grown in several culture media with increasing test
concentrations of the compound, and the minimal concentration of
the compound yielding a mass-spectrometry spectrum detectably
different from the mass-spectrometry spectrum obtained from the
cell-extract of the microorganism strain grown in the compound-free
culture medium is determined.
[0030] As intended herein the "minimal concentration of the
compound yielding a mass-spectrometry spectrum significantly
different from the mass-spectrometry spectrum obtained from the
cell-extract of the microorganism strain grown in the compound-free
culture medium" is also called the minimal profile (i.e. mass
spectrometry spectrum) change concentration (MPCC). Advantageously,
the inventors have shown that for a given cell strain and compound
the MPCC and the MIC are correlated.
[0031] Preferably, determining the minimal concentration
comprises:
[0032] obtaining a mass spectrometry spectrum from a protein
extract of the cell strain grown in the culture medium having the
highest test concentration of the several culture media;
[0033] comparing the mass spectrometry spectra obtained from
protein extracts of the several culture media to (i) the mass
spectrometry spectrum from a protein extract of the cell strain
grown in the culture medium having the highest test concentration
and (ii) the mass spectrometry spectrum from the compound-free
culture medium;
[0034] selecting the mass spectrometry spectrum obtained from a
protein extract of the several culture media with the lowest
compound concentration and which presents more similarity with the
mass spectrometry spectrum (i) than with the mass spectrometry
spectrum (ii);
the lowest compound concentration being the minimal concentration
to be determined.
[0035] Determining that a mass spectrometry spectrum presents more
similarity, or shares more resemblance, with a first spectrum than
with a second spectrum can be routinely determined by one of skill
in the art, in particular using a similarity measure based on
Spearman rank correlation coefficients.
[0036] More particularly, once the mass spectrometry spectra have
been obtained from protein extracts of the several culture media, a
mean spectrum can be determined. Then simple peak detection can be
performed on this mean spectrum, and the final peak locations can
be selected based on the mean intensity of the peak.
[0037] In a first step of statistical analysis, testing whether
there is a difference between the extreme concentration spectra may
be achieved. This may can be routinely determined by one of skill
in the art performing an exact permutation test using Spearman rank
correlation coefficient as a similar measure. Briefly, all the rank
correlation coefficients between all the spectra may first be
calculated. The mean of the intra-class rank correlations
coefficients (IntraRCCM) and the mean of the inter-class rank
correlation coefficients (InterRCCM) may then be determined.
[0038] The test criterion may be the ratio InterRCCM/IntraRCCM
under the null hypothesis of no difference between class
memberships, the criterion's expected value being 1. When class
memberships are informative, interRCCM is lower than intraRCCM, and
expected criterion values are lower than 1. The permutation test
can be achieved by computing the distribution of the criterion for
all the permutations of the class memberships.
[0039] Once the difference between extreme concentrations has been
statistically proved, the minimal concentration at which a
particular spectrum starts to differ significantly from the null
control spectrum one may be determined. This can be achieved by
computing for each concentration, the corresponding spectrum
similarity with each spectrum from the two extreme concentrations,
and by classifying it as "near of the null concentration" or "near
of the maximum concentration" according to the similarity values,
using the mean inter-class rank correlation coefficient
(InterRCCM). The minimal profile change concentration (MPCC) is
defined as the minimum concentration that is more similar to the
maximum concentration than to the null one.
[0040] In a preferred embodiment, the invention relates to a device
for implementing the above-defined method in which the cell strain
is grown in several culture media with increasing test
concentrations of the compound, and the minimal concentration of
the compound yielding a mass-spectrometry spectrum detectably
different from the mass-spectrometry spectrum obtained from the
cell-extract of the microorganism strain grown in the compound-free
culture medium is determined.
[0041] Such device comprises;
[0042] means for obtaining a mass-spectrometry spectrum for a
protein extract of the yeast strain grown in a first compound-free
culture medium and in several culture media with increasing test
concentrations of the compound;
[0043] means for comparing the mass spectrometry spectra obtained
for protein extracts of the several culture media with increasing
test concentrations of the compound to (i) the mass spectrometry
spectrum of a protein extract of the yeast strain grown in the
culture medium having the highest assayed test concentration and
(ii) the mass spectrometry spectrum of a protein extract of the
yeast strain grown in the compound-free culture medium;
[0044] means for determining the minimal concentration by
determining the mass spectrometry spectrum obtained from a protein
extract of the several culture media with the lowest compound
concentration and which presents more similarity with the mass
spectrometry spectrum (i) than with the mass spectrometry spectrum
(ii); the lowest compound concentration being the minimal
concentration.
[0045] A device 1 for performing the data analysis is schematically
illustrated in FIG. 5. Device 1 comprises processing means, such as
a Central Processing Unit 2, storage means, such as a Random Access
or Read-Only memory 4 and a database 6, human-machine interface
means, such as a Liquid Crystal Display 8 together with a keyboard
10, and an Input/Output interface, such as an RS 232 connection
12.
[0046] The method according to the invention is realised by means
of a software, the instructions of which are stored in memory 4 and
are processed by CPU 2.
[0047] In a first step, data acquisition is performed. The mass
spectrometer MS is plugged onto the Input/Output interface 12. The
data corresponding to the spectrum obtained from a sample currently
analysed with the mass spectrometer MS are transferred to device
1.
[0048] Once the transfer of a spectrum is completed, the user
labels it with the value of the FCZ concentration of the class of
the sample, and an integer between 1 and n to identify said sample
in said class.
[0049] The spectrum is then stored in database 6 with an Id
corresponding to said concentration and said integer.
[0050] After the n spectra for the c values of the FCZ
concentration are acquired, device 1 is put in a pre-processing
mode by the user.
[0051] In this pre-processing mode, peak extraction algorithm is
performed. The average of the n.times.c spectrum is calculated. The
base line of the average spectrum is determined. Finally, on the
average spectrum, peaks are retained for a signal to noise ratio
greater than 4 times the value of the base line. Only the peaks
with a m/z ratio between 3000 and 20000 are retained for the
statistical analysis. This leads to the extraction of a set of N
characteristic m/z ratios where peaks occur.
[0052] The pre-processing algorithm then comprises the
discretisation of each of the n.times.c spectra in database 6. A
discretised spectrum is derived from each spectrum by reading the
values of the m/z ratio for the N characteristic m/z ratios where
peaks occur.
[0053] Finally, the pre-processing algorithm runs a rank list
creation routine for associating a ranked spectrum to each
discretised spectrum. Each coordinate of the discretised spectrum
is replaced by its rank in the ordered list of the N m/z ratios of
said discretised spectrum.
[0054] Each thus obtained ranked spectrum is stored in database 6
with the Id of the corresponding initial spectrum.
[0055] Then, device 1 is put in a statistical analysis mode where
CPU 2 processes the following comparison algorithm:
[0056] The first step consists in selecting two classes of n
spectra. These two classes are respectively the class corresponding
to a null value of the FCZ concentration and the class
corresponding to the maximum FCZ concentration. The corresponding
2n ranked spectra are retrieved from database 6 and stored into
memory 4.
[0057] CPU 2 then calculates the n(2n-1) correlation rank
coefficients, one coefficient for each possible pair of ranked
spectra. The average of the correlation rank coefficients of the
pairs of spectra from the same class (same concentration of FCZ)
leads to the determination of the IntraRCCM parameter. The average
of the correlation rank coefficients of the pairs of specta from
different classes (different concentrations of FCZ) leads to the
determination of the InterRCCM parameter.
[0058] Finally, CPU 2 computes the ratio interRCCM/IntraRCCM by
dividing the InterRCCM parameter by the IntraRCCM parameter before
comparing it with unity.
[0059] If the ratio interRCCM/IntraRCCM is equal to unity, it means
that there is no difference between the two classes and the
statistical analysis ends.
[0060] On the other hand, if ratio interRCCM/IntraRCCM is lower
than 1 it means that there is a difference between the two classes
with extreme concentration.
[0061] In this case, CPU 2 processes the following similarity
algorithm:
[0062] For a class corresponding to an intermediary value of the
FCZ concentration, a null interRCCM parameter is calculated between
the intermediary class and the null concentration class and a
maximum interRCCM parameter is calculated between the intermediary
class and the maximum concentration class.
[0063] The intermediary class is said more similar to the maximum
concentration class than to the null concentration class when the
maximum interRCCM parameter is greater than the null interRCCM
parameter.
[0064] At the end, the minimal profile change concentration MPCC is
given by the smallest of the concentrations of the intermediary
classes that are more similar to the maximum concentration
class.
[0065] This MPCC value can be displayed on the LCD screen at the
end of the processing. It is stored in database 6.
[0066] As a variant, other statistical analysis methods may be
used.
[0067] The invention will be further described by the following
non-limiting figures and Examples.
DESCRIPTION OF THE FIGURES
[0068] FIG. 1 represents alterations in the mass spectra of the
DSY2260 C. albicans strain exposed to increasing fluconazole (FCZ)
concentrations (virtual gel). The x-axis represents m/z value, on
the left the y-axis shows running spectrum number, while on the
right peak intensity is expressed in a grey colour scale with
arbitrary units (a.u.). Briefly, 10.sup.6 yeasts/ml were cultured
for 15 h, with three biological replicates for each FCZ
concentration (from 128 to 0.125 .mu.g/ml) and for unexposed
yeasts. Acidic extracts were analysed in duplicate by MALDI-TOF
MS.
[0069] FIGS. 2 and 3 represent the average mass spectra of the 6
replicate spectra of DSY2260 C. albicans exposed to 2 .mu.g/ml of
FCZ (FIG. 2) and 4 .mu.g/ml of FCZ (FIG. 3) (mass range 5800-7600
m/z).
[0070] FIG. 4 represents the correlation between MICs evaluated by
the CLSI methodology and MPCCs determined by a MALDI-TOF MS method
according to the invention (regression line).
[0071] FIG. 5 is a schematic representation of the apparatus to
perform an analysis method according to the invention.
EXAMPLE
1. Methods
Yeasts Strains
[0072] All experiments were performed using sequential Candida
albicans clinical isolates. Eight groups of sequential related
isolates (groups A to K) were characterized. All strains were
isolated from HIV-positive patients with oropharyngeal candidasis
that were treated mainly with FCZ. Each group contains one azole
susceptible strain (FCZ MIC ranged from 0.125 to 8 .mu.g/ml) and
its related-sequential strains with higher FCZ MICs (ranging from
16 to 128 .mu.g/ml) within the same MLST genotype. For each
clinical isolate, MICs according to CLSI broth-microdilution
methodology 1, and presence of ERG11 and TAC mutations, and/or
CDR1/2 and MDR hyperexpression were determined previously 2 (Table
1). The ATCC 90028 C. albicans strain (MIC=0.25 .mu.g/ml) was added
as a reference strain.
Antifungal Agents
[0073] Fluconazole (FCZ) pure powder (Sigma Chemical CO.,
Saint-Louis, Mo., USA) was dissolved in pure water. Serial
dilutions of drug (concentration ranged from 256 to 0.25 .mu.g/ml),
made into RPMI 1640 medium (with glutamine and without bicarbonate,
Invitrogen) buffered with MOPS (0.165M) (Sigma Chemical CO.,
Saint-Louis, Mo., USA) and adjusted to pH 7 with sodium hydroxide
(5N), were dispensed in 600 .mu.l aliquots into sterile 24-well
flat-bottomed microtiter plates.
Samples Preparation
[0074] After cultivation during 48 h at 37.degree. C. on Sabouraud
agar medium, yeasts cells were transferred into RPMI 1640 medium
(with glutamine but without bicarbonate, Invitrogen) buffered with
MOPS (0.165M) (Sigma Chemical CO., Saint-Louis, Mo., USA), and
adjusted to pH 7 with sodium hydroxide (5N). Then, 600 .mu.l of
RPMI with yeast (2.106 yeasts/imp were added to 600 .mu.l of RPMI
with FCZ into microtiter plates (final FCZ concentration ranged
from 128 to 0.125 .mu.g/ml), or in RPMI alone as negative control.
Culture was performed in a 30.degree. C. incubator for 15 hours,
with continuous agitation.
[0075] Yeast extraction was performed as follows: [0076] (1)
samples were centrifuged and the supernatant removed before the
pellet was washed twice in pure water; [0077] (2) the pellet was
then resuspended in 300 .mu.L of pure water, and 900 .mu.L of
ethanol were then added; [0078] (3) after another round of
centrifugation, 50 .mu.L of 70% formic acid and 50 .mu.L pure
acetonitrile were added to the residual pellet and subsequent
solution was repeatedly and thoroughly vortexed before a final
centrifugation. Each centrifugation was performed at 10 000 rpm for
10 minutes at room temperature.
MALDI-TOF Mass Spectrometry
[0079] The supernatant was distributed (0.5 .mu.l droplet) in
duplicate on a MALDI AnchorChip sample slide (Bruker-Daltonics,
Bremen, Germany), then air-dried. The
.alpha.-cyano-4-hydroxycinnamic acid (CHCA) matrix
(Bruker-Daltonics), prepared at a concentration of 50 mg/ml in 50%
acetonitrile and 50% water with 0.1% TFA, was sonicated for 5 min
before being spotted (0.5 .mu.l) over the dried sample. A DH5a
Escherichia coil protein extract (Bruker-Daltonics) was deposited
on the calibration spot of the Anchorchip for external calibration.
MALDI analysis were performed on a Bruker Autoflex I MALDI TOF mass
spectrometer with a nitrogen laser (337 nm) operating in linear
mode with delayed extraction (260 ns) at 20 kV accelerating
voltage. Each spectrum was automatically collected in the positive
ion mode as an average of 500 laser shots (50 laser shots at 10
different spot positions). Laser energy was set just above the
threshold for ion production. Mass range between 3,000 and 20,000
m/z (ratio mass/charge) was selected to collect the signals with
the AutoXecute tool of flexControl acquisition software (Version
2.4; Bruker-Daltonics). Only peaks with a signal/noise ratio >3
were considered. Spectra were eligible for further analysis when
the peaks had a resolution better than 600. For each cultivation
condition, we collected mass spectra from 3 biological replicates
and 2 technical replicates.
Statistical Analysis
a) Pre-Processing Ending
[0080] Data obtained from the flex control acquisition software
were a set of lists of mass/charge peaks and corresponding
intensities, one list for each spectrum. For each FCZ concentration
(11 FCZ concentrations and without FCZ), 6 spectra, obtained from 3
biological replicates tested in duplicate, were collected, so the
whole initial set of data consisted of 72 spectra.
[0081] Alignment was performed using a method based on the mean
spectrum, using an approach similar to that recommended by Coombes
et al. (Coombes K R, Baggerly K A, and Morris J S: Pre-Processing
Mass Spectrometry Data. Fundamentals of Data Mining in Genomics and
Proteomics, W Dubitzky, M Granzow, and D Berrar, eds. Kluwer,
Boston, 79-99 (2007)).
[0082] The mean spectrum of the 72 spots was computed, a simple
peak detection algorithm was run on the mean spectrum, and a
selection of the final peak locations was based on the mean
intensity of the peak. For each spectrum, the peaks corresponding
to the final peak locations were retained for the statistical
analysis. An error of 0.05% of the mass/charge ratio was allowed in
the two last steps.
b) Comparison Between Extreme Concentrations
[0083] At this stage, data consisted of 12 ordered subsets of 6
spectra, each subset corresponding to a concentration value, from 0
to 128 .mu.g/ml. The first step of statistical analysis was aimed
at testing whether there was a difference between the extreme
concentration spectra.
[0084] An exact permutation test using Spearman rank correlation
coefficient as a similar measure, based on the following
computations, has been carried out: [0085] in a first step, all the
rank correlation coefficients between the 12 spectra were computed;
[0086] in a second step, the mean of the intra-class rank
correlations coefficients (IntraRCCM) and the mean of the
inter-class rank correlation coefficients (InterRCCM) were
computed.
[0087] The test criterion is the ratio InterRCCM/IntraRCCM Under
the null hypothesis of no difference between class memberships, the
criterion's expected value is 1. When class memberships are
informative, interRCCM is lower than intraRCCM, and expected
criterion values are lower than 1. The permutation test is achieved
by computing the distribution of the criterion for all the
permutations of the class memberships, and the exact p-value is the
proportion of criteria lower or equal to the one corresponding to
the observed criterion value.
c) Determination of the MPCC
[0088] Once the difference between extreme concentrations has been
statistically proved, the aim of the next step is to find the
minimal concentration at which a particular spectrum starts to
differ significantly from the null control spectrum one. This is
achieved by computing for each concentration, the corresponding
spectrum similarity with each from the two extreme concentrations,
and by classifying it as "near of the null concentration" or "near
of the maximum concentration" according to the similarity values.
The similarity function used is the mean inter-class rank
correlation coefficient (InterRCCM). The minimal profile change
concentration (MPCC) is defined as the minimum concentration that
is more similar to the maximum concentration than to the null
one.
2. Results
[0089] In order to obtain standardized fingerprints, the inventors
first optimized the protocols. All experiments were carried out
using the C. albicans susceptible reference strain, ATCC 90028
(Methods). The influence of starting inoculum, the concentration of
yeast cells in the sample to be analyzed and thus the minimum
culturing time required, on the quality and reproducibility of the
fingerprint reproducibility were assessed. This allowed
establishing that optimal results are obtained from cultures
initiated with 10.sup.6 yeasts/ml.
[0090] The inventors then determined the effect that varying
concentrations of FCZ (serial dilution from 128 to 0.125 .mu.g/ml)
would have on C. albicans mass spectrometry fingerprint patterns
(Methods).
[0091] The yeast cell pellets obtained from cultures grown for 15 h
were subjected to acid extraction and the supernatants analyzed by
MALDI-TOF MS. A typical result is presented in FIG. 1 where the
mass spectrometry spectra of the DSY2260 clinical C. albicans
strain (FCZ MC=8 .mu.g/ml) from cultures exposed to serial FCZ
dilutions are shown. For this strain, the mass spectrometry spectra
from cultures exposed up to 2 .mu.g/ml FCZ are indistinguishable
from those observed for the control culture. Conversely, the
spectra deviate from the control in a detectable and quantifiable
manner (through gain or loss of spectra) in the cultures exposed to
4 .mu.g/ml and upwards (FIGS. 2 and 3).
[0092] Accordingly, these observations led the inventors to
formulate a new endpoint, namely the minimal profile change
concentration (MPCC), a value defined as the lowest drug (FCZ in
the experiment above) concentration at which a mass spectrum
profile change can be detected.
[0093] Besides, the inventors have further devised a novel
statistical approach to calculate this value objectively (Methods),
which is based on the mass and intensity of each peak in the
fingerprints. In this statistical analysis, the discrepancies
between the mass spectra at the two extreme conditions (128
.mu.g/ml FCZ and no FCZ) are first defined. Then, the similarity to
each of the two "extreme" spectra is statistically evaluated for
the spectrum recorded at each of the different intermediate FCZ
concentrations, to yield a classification of "nearer to the 128
.mu.g/ml" or "nearer to the FCZ negative" spectrum.
[0094] The validity of the new methodology above was established as
follows. The MPCC of the reference strain (C. albicans ATCC 90028,
MIC=0.25 .mu.g/ml) was determined, as well as that of sixteen C.
albicans isolates with distinct drug resistance profiles. Eight are
known to have low-MICs (range 0.125 to 8 .mu.g/ml) and the other
eight have high-MICs (range 16 to 128 .mu.g/ml) to FCZ, these MICs
having been determined by the CLSI standard method. In addition,
the FCZ-resistant strains tested are representative of the
different resistance mechanisms, mating types and clades 15 that
are found in C. albicans (Table 1).
[0095] Comparison of MPCCs with the MICs for the 17 strains above
(FIG. 4 and Table 1) revealed a very high degree of concordance,
with full agreement for 94% and 100% of the strains, depending on
whether concordance was scored as +/-1 or within +/-2 drug dilution
values, respectively. A cut-off set at +/-2 drug dilution values is
the maximum accepted discrepancy for agreement in comparison
conducted between antifungal susceptibility testing methods
(Espinel-Ingroff et al. J. Clin Microbiol 43, 3884-3889
(2005)).
[0096] Importantly, the MPCC determinations were concordant and
accurate irrespective of the type of drug resistance mechanism
(ERG11 mutations, TAG mutation, CDR1/2 or MDR hyper-expression),
the mating type, or the clade to which the different strains tested
belonged. These results not only validate the new methodology, but
also offer strong indication of the robustness of the methodology
present in the face of C. albicans strains with diverse genetic
backgrounds.
[0097] As it stands, where a mass spectrometer is already
available, the running costs for each sample analyzed by the method
presented here is slightly less than 1 Euro, which compares quite
favourably with the higher costs of all other methods. The
diagnostic profile shift observed for FCZ does not vary with either
the genetic background of the strain tested, nor the level or
mechanism of the resistance to FCZ. Accordingly, it is likely that
a characteristic profile would be associated with each of the
different classes of drugs known to inhibit a given pathogen (e.g.
triazoles echinocandins, polyene, and antimetabolites for C.
albicans). The present methodology appears to be suitable for
monitoring the emergence of resistance to drugs used against
pathogenic organisms, including bacteria and eukaryotic cancer
cells or pathogens such as Cryptosporidiurn and Plasmodium.
TABLE-US-00001 TABLE 1 Description of C. albicans strains used in
this study. FCZ FCZ C. albicans MICs MPCCs CDR1/ TAC1 ERG11 strains
Name Group (.mu.g/ml) (.mu.g/ml) CDR2 MDR1 MTL mutations mutations
Clade Reference ATCC 0.25 0.25 na MEB strain 90028 Clinical DSY281
A 1 0.5 a/.alpha. S450F* 1 isolates DSY284 32 8 X a/.alpha. G980E*
S450F 1 (grouped by DSY 294 B 0.25 0.25 a/.alpha. G129A na
sequential DSY296 128 128 X .alpha./.alpha. N977D G464S, na
isolates G129A from DSY544 C 0.5 0.25 a/.alpha. D116E*, 1 selected
DYS775 K128T* patients) 128 64 X a/a G980W G464S 1 DSY2260 D 8 4 X
a/.alpha. E266D, na DSY2262 G464S 64 32 X a/.alpha. E266D, na
L376V, G464S DSY2321 E 0.25 0.5 a/.alpha. S405F* na DSY2322 16 16 X
.alpha./.alpha. G980E S405F na DSY347 F 0.25 0.125 a/.alpha. na
DSY289 128 128 X a/.alpha. A736V S405F, na Y132H DSY290 G 0.5 0.25
a/.alpha. S DSY292 128 128 X a/.alpha. G464S, S R467K, Y132H*
DSY2284 H 0.25 0.5 a/.alpha. 11 DSY2285 16 16 X a/a/a/.alpha. 11
Resistance mechanisms, mating type, and clade were previously
determined. Agreement between MICs to MPCCs was independent from
the genetic background. *Heterozygous mutation, S: Singleton, na:
Not available
* * * * *